Codes of Culture | Issue 89
Anthropic sues, LeCun bets big on AI startup, and AI glasses are here
Welcome back to Codes of Culture. I’m Ashumi Sanghvi.
This week, three stories broke within hours of each other on Sunday night, and I found myself reading them in sequence, struck by how much they were saying the same thing in different registers. A lawsuit. A billion-dollar seed round with no product. Europe’s largest-ever Series C for a company most people outside the industry have never heard of.
Each one is a different angle on the same underlying question: who controls the infrastructure of AI, and what are they willing to fight for to keep it? The hardware stories that follow ask something adjacent: what does it actually feel like to wear that future on your body? A lot moved this week.
Scroll on for 5 stories in this mid-week issue, which I think deserve the most attention.
Remember to subscribe for full access to our news, insights reports, and global events. We are currently working on being at F1 in Miami, the Milken Global Conference in LA, and Cannes Lions. Reply to this email if you will be there too and want to stay in the loop for what we get up to.
📖In this issue:
Anthropic sued the Pentagon. Every AI contract is now in scope.
LeCun bets $1B that LLMs are a dead end
Europe’s biggest-ever AI Series C — and why it is a compute story.
Samsung reveals its new AI smart glasses
The AI ring that wants to replace your notebook. Sandbar closes $23M.
Anthropic sued the Pentagon. Every AI contract is now in scope.
What’s happening: Anthropic filed two federal lawsuits against the US DOD on 9 March 2026 after the Pentagon designated the company a supply chain risk, a classification historically reserved for foreign adversaries. The dispute centres on Anthropic’s refusal to permit Claude for mass surveillance of Americans or fully autonomous weapons without human oversight. The DOD sought unrestricted access for all lawful purposes. OpenAI reached a deal with the Pentagon within hours. More than 30 OpenAI and Google DeepMind employees, including DeepMind chief scientist Jeff Dean, filed an amicus brief in Anthropic’s support. Claude briefly surpassed ChatGPT in the US App Store following the designation.
TLDR:
Anthropic refused the DOD’s request regarding autonomous weapons and domestic surveillance. The Pentagon responded with a supply chain risk designation.
That label is historically reserved for foreign adversaries. Applying to a US AI company for a product position is unprecedented.
OpenAI reached a deal with the DOD the same day. 30+ employees from OpenAI and Google DeepMind filed in support of Anthropic.
The case turns on one question: can product guardrails be treated as a national security threat?
Every founder or operator with government adjacency needs to be tracking this. The precedent will not stop at Anthropic.
Why it matters: This is a commercial and constitutional-precedent story, not an AI safety story. Anthropic’s position: product limits cannot justify a national security designation. The Pentagon’s counter: the military cannot be subject to contractor-imposed constraints on lawful use. For founders and operators in our network building AI products with any government adjacency, this case will set the frame for every enterprise AI contract negotiation that follows.
LeCun bets $1B that LLMs are a dead end.
What’s happening: Months after leaving Meta, Turing Prize winner Yann LeCun closed a $1.03B seed round for AMI Labs at a $3.5B pre-money valuation. The startup is less than three months old, with no product and no revenue. The thesis is confrontational: LLMs are not the end state, and the real frontier lies in world models, AI systems built to reason about physical reality rather than predict the next word. Bezos Expeditions, Temasek, Nvidia, Eric Schmidt, and others backed the round. The first commercial partner is Nabla, the digital health startup co-founded by AMI Labs CEO Alexandre LeBrun. HQ: Paris.
TLDR:
One of the largest seed financings ever for a company with no product. The bet is entirely on the thesis.
World models: AI with memory, spatial reasoning, and planning capability, versus LLMs that pattern-match language.
Paris HQ is deliberate—European AI positioning as an alternative architecture, not just an alternative geography.
Bezos Expeditions, Temasek, Nvidia, and Eric Schmidt among backers. Conviction capital, not a speculative swing.
Downstream implications for luxury, healthcare, and industrial automation are significant if the thesis holds.
Why it matters: If world models mature, AI shifts from generating outputs to genuinely operating in environments, with memory, planning, and physical consequences. The $1B without a product is the tell: this is a paradigm bet. Future+ has been tracking European AI as a long-term structural story, and AMI Labs is the clearest signal yet that Paris intends to be a credible alternative to the Silicon Valley consensus on architecture.
Europe has a new decacorn, and it’s a compute company, not a model company.
What’s happening: Nscale closed the largest Series C in European history: $2B at a $14.6B valuation, co-led by Aker and 8090 Industries with Goldman Sachs, JPMorgan, Dell, Nvidia, Citadel, and Jane Street participating. Former Meta COO Sheryl Sandberg, former Yahoo president Susan Decker, and former UK Deputy Prime Minister Nick Clegg join the board. Nscale’s Stargate Norway JV targets 100,000 Nvidia GPUs by the end of 2026 with OpenAI as anchor customer. Stargate UK adds a second sovereign compute node. CEO Josh Payne has signalled an IPO as early as this year.
TLDR:
A European decacorn built on physical infrastructure, not a model. The moat is vertical integration across energy, data centres, compute, and orchestration.
Sandberg, Clegg, and Decker on the board: political fluency, regulatory navigation, and public-market experience ahead of a likely IPO.
Stargate Norway: 100,000 Nvidia GPUs by year-end, OpenAI as anchor. Sovereign compute is being built now.
Goldman Sachs, JPMorgan, Citadel, Jane Street in the round. This is financial infrastructure capital, not pure venture.
Europe’s AI story is increasingly being written in hardware and governance. This raise is one of the clearest expressions of that.
Why it matters: Sovereign compute is the story underneath the European AI story. Nscale is building control over the physical substrate on which everything else runs. Sandberg and Clegg bring the consumer, policy, and regulatory muscle that a company navigating EU AI governance and a public listing will need. For Future+ clients operating across Europe and the Middle East: as compute becomes sovereign, infrastructure relationships are what matter.
Samsung reveals its new AI smart glasses.
What’s happening: At Mobile World Congress in Barcelona, Samsung EVP Jay Kim gave the first substantive public briefing on the company’s AI smart glasses, confirming a 2026 launch. Kim confirmed the device will carry a camera positioned at eye level and be connected to a Galaxy smartphone, which handles all AI processing. The glasses feed what the wearer sees to the phone; the phone processes it and returns information. Kim declined to confirm whether the glasses will include a built-in display, pointing to Samsung’s existing watch and phone lineup for screen needs. Qualcomm CEO Cristiano Amon separately confirmed the 2026 timeline, framing the glasses as the next frontier for agentic AI. Meta’s Ray-Ban glasses currently hold 82% of the global smart glasses market, according to Counterpoint Research.
TLDR:
Eye-level camera, smartphone tethering, Galaxy AI. The glasses are sensors; the phone is the brain.
No confirmed built-in display in V1. Kim’s framing: if you need a screen, use your phone or watch.
Qualcomm confirmed 2026 alongside Samsung. The Qualcomm-Google-Samsung partnership on XR has been running since 2023.
Meta holds 82% market share. Samsung is entering the category, not yet disrupting it; the first product is described as targeting the industry.
Agentic use cases are the pitch: glasses that see your environment and act on your behalf, booking rides, surfacing information, without you touching a phone.
Why it matters: The strategic shift here is not the hardware. It is the input model. Glasses that continuously see what you see create a fundamentally different data layer than a phone you occasionally raise. For luxury and creative brands in our network: the person wearing your product will soon be wearing a camera at eye level. The contextual and proximity brand moment that is created is not a future consideration. It is a current design brief.
The AI ring that wants to replace your notebook just raised $23M.
What’s happening: Sandbar, founded by former Meta employees Mina Fahmi and Kirak Hong, closed a $23M Series A led by Adjacent and Kindred Ventures, bringing total funding to $36M. The Stream ring features a proximity-tuned microphone that activates when you hold the touch panel and lift your hand to your face, capturing voice notes, AI assistant queries, and media controls. Pre-orders for the first batch sold out; a second batch followed. Sandbar plans to begin shipping this summer. The founding team previously worked at CTRL-Labs, Magic Leap, Amazon, Fitbit, and Google.
TLDR:
Stream is a smart ring for note-taking, not health tracking. Microphone off by default, activated by gesture. Proximity-tuned to capture you, not the room.
First pre-order batch sold out. Some early users are logging over 50 interactions a day for planning, trips, and presentations.
Shipping this summer. Roadmap includes agentic workflows, so captured notes can trigger actions, not just transcriptions.
A design-forward wearable aimed at broad adoption, not just tech early adopters.
The note-taking hardware category is growing: Plaud, Pebble, Taya, and now Sandbar. The form factor race is on.
Why it matters: Sandbar sits at the intersection of two things Future+ tracks closely: ambient wearable AI and the premium hardware design question. The gesture activation — lifting your hand to your face to record — makes private use legible in public. Wearables that capture ambient context are quietly building the data layer for the next generation of personal AI. The ring is the form factor that is actually wearable.
→ Codes of Culture is published by Future+, a global network and strategic advisory firm, founded by Ashumi Sanghvi. To explore the Future+ network and strategic partnerships, click here. To subscribe, access the full archive, or get in touch: codesofculture.futureplus.xyz.
Disclaimer: Information in this newsletter is only for the purpose of sharing knowledge, news and insights.
 | A guest post by
|